_ Git - youtube-dl/blob - youtube_dl/extractor/youtube.py

   1 # coding: utf-8
   2
   3 import json
   4 import netrc
   5 import re
   6 import socket
   7 import itertools
   8
   9 from .common import InfoExtractor, SearchInfoExtractor
  10 from ..utils import (
  11     compat_http_client,
  12     compat_parse_qs,
  13     compat_urllib_error,
  14     compat_urllib_parse,
  15     compat_urllib_request,
  16     compat_str,
  17
  18     clean_html,
  19     get_element_by_id,
  20     ExtractorError,
  21     unescapeHTML,
  22     unified_strdate,
  23     orderedSet,
  24 )
  25
  26 class YoutubeBaseInfoExtractor(InfoExtractor):
  27     """Provide base functions for Youtube extractors"""
  28     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
  29     _LANG_URL = r'https://www.youtube.com/?hl=en&persist_hl=1&gl=US&persist_gl=1&opt_out_ackd=1'
  30     _AGE_URL = 'http://www.youtube.com/verify_age?next_url=/&gl=US&hl=en'
  31     _NETRC_MACHINE = 'youtube'
  32     # If True it will raise an error if no login info is provided
  33     _LOGIN_REQUIRED = False
  34
  35     def report_lang(self):
  36         """Report attempt to set language."""
  37         self.to_screen(u'Setting language')
  38
  39     def _set_language(self):
  40         request = compat_urllib_request.Request(self._LANG_URL)
  41         try:
  42             self.report_lang()
  43             compat_urllib_request.urlopen(request).read()
  44         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
  45             self._downloader.report_warning(u'unable to set language: %s' % compat_str(err))
  46             return False
  47         return True
  48
  49     def _login(self):
  50         (username, password) = self._get_login_info()
  51         # No authentication to be performed
  52         if username is None:
  53             if self._LOGIN_REQUIRED:
  54                 raise ExtractorError(u'No login info available, needed for using %s.' % self.IE_NAME, expected=True)
  55             return False
  56
  57         request = compat_urllib_request.Request(self._LOGIN_URL)
  58         try:
  59             login_page = compat_urllib_request.urlopen(request).read().decode('utf-8')
  60         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
  61             self._downloader.report_warning(u'unable to fetch login page: %s' % compat_str(err))
  62             return False
  63
  64         galx = None
  65         dsh = None
  66         match = re.search(re.compile(r'<input.+?name="GALX".+?value="(.+?)"', re.DOTALL), login_page)
  67         if match:
  68           galx = match.group(1)
  69         match = re.search(re.compile(r'<input.+?name="dsh".+?value="(.+?)"', re.DOTALL), login_page)
  70         if match:
  71           dsh = match.group(1)
  72
  73         # Log in
  74         login_form_strs = {
  75                 u'continue': u'https://www.youtube.com/signin?action_handle_signin=true&feature=sign_in_button&hl=en_US&nomobiletemp=1',
  76                 u'Email': username,
  77                 u'GALX': galx,
  78                 u'Passwd': password,
  79                 u'PersistentCookie': u'yes',
  80                 u'_utf8': u'霱',
  81                 u'bgresponse': u'js_disabled',
  82                 u'checkConnection': u'',
  83                 u'checkedDomains': u'youtube',
  84                 u'dnConn': u'',
  85                 u'dsh': dsh,
  86                 u'pstMsg': u'0',
  87                 u'rmShown': u'1',
  88                 u'secTok': u'',
  89                 u'signIn': u'Sign in',
  90                 u'timeStmp': u'',
  91                 u'service': u'youtube',
  92                 u'uilel': u'3',
  93                 u'hl': u'en_US',
  94         }
  95         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
  96         # chokes on unicode
  97         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k,v in login_form_strs.items())
  98         login_data = compat_urllib_parse.urlencode(login_form).encode('ascii')
  99         request = compat_urllib_request.Request(self._LOGIN_URL, login_data)
 100         try:
 101             self.report_login()
 102             login_results = compat_urllib_request.urlopen(request).read().decode('utf-8')
 103             if re.search(r'(?i)<form[^>]* id="gaia_loginform"', login_results) is not None:
 104                 self._downloader.report_warning(u'unable to log in: bad username or password')
 105                 return False
 106         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 107             self._downloader.report_warning(u'unable to log in: %s' % compat_str(err))
 108             return False
 109         return True
 110
 111     def _confirm_age(self):
 112         age_form = {
 113                 'next_url':     '/',
 114                 'action_confirm':   'Confirm',
 115                 }
 116         request = compat_urllib_request.Request(self._AGE_URL, compat_urllib_parse.urlencode(age_form))
 117         try:
 118             self.report_age_confirmation()
 119             compat_urllib_request.urlopen(request).read().decode('utf-8')
 120         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 121             raise ExtractorError(u'Unable to confirm age: %s' % compat_str(err))
 122         return True
 123
 124     def _real_initialize(self):
 125         if self._downloader is None:
 126             return
 127         if not self._set_language():
 128             return
 129         if not self._login():
 130             return
 131         self._confirm_age()
 132
 133 class YoutubeIE(YoutubeBaseInfoExtractor):
 134     IE_DESC = u'YouTube.com'
 135     _VALID_URL = r"""^
 136                      (
 137                          (?:https?://)?                                       # http(s):// (optional)
 138                          (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/|
 139                             tube\.majestyc\.net/)                             # the various hostnames, with wildcard subdomains
 140                          (?:.*?\#/)?                                          # handle anchor (#/) redirect urls
 141                          (?:                                                  # the various things that can precede the ID:
 142                              (?:(?:v|embed|e)/)                               # v/ or embed/ or e/
 143                              |(?:                                             # or the v= param in all its forms
 144                                  (?:(?:watch|movie)(?:_popup)?(?:\.php)?)?    # preceding watch(_popup|.php) or nothing (like /?v=xxxx)
 145                                  (?:\?|\#!?)                                  # the params delimiter ? or # or #!
 146                                  (?:.*?&)?                                    # any other preceding param (like /?s=tuff&v=xxxx)
 147                                  v=
 148                              )
 149                          )?                                                   # optional -> youtube.com/xxxx is OK
 150                      )?                                                       # all until now is optional -> you can pass the naked ID
 151                      ([0-9A-Za-z_-]+)                                         # here is it! the YouTube video ID
 152                      (?(1).+)?                                                # if we found the ID, everything can follow
 153                      $"""
 154     _NEXT_URL_RE = r'[\?&]next_url=([^&]+)'
 155     # Listed in order of quality
 156     _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13',
 157                           '95', '94', '93', '92', '132', '151',
 158                           '85', '84', '102', '83', '101', '82', '100',        # 3D
 159                           '138', '137', '136', '135', '134', '133', '160',    # Dash video mp4
 160                           '141', '140', '139',                                # Dash auido mp4
 161                           '248', '247', '246', '245', '244', '243', '242',    # Dash video webm
 162                           '172', '171',                                       # Dash audio webm
 163                           ]
 164     _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13',
 165                                       '95', '94', '93', '92', '132', '151',
 166                                       '85', '102', '84', '101', '83', '100', '82',
 167                                       '248', '247', '246', '245', '244', '243', '242',    # Dash video webm
 168                                       '172', '171',                                       # Dash audio webm
 169                                       '138', '137', '136', '135', '134', '133', '160',    # Dash video mp4
 170                                       '141', '140', '139',                                # Dash auido mp4
 171                                       ]
 172     _video_extensions = {
 173         '13': '3gp',
 174         '17': 'mp4',
 175         '18': 'mp4',
 176         '22': 'mp4',
 177         '37': 'mp4',
 178         '38': 'mp4',
 179         '43': 'webm',
 180         '44': 'webm',
 181         '45': 'webm',
 182         '46': 'webm',
 183
 184         # 3d videos
 185         '82': 'mp4',
 186         '83': 'mp4',
 187         '84': 'mp4',
 188         '85': 'mp4',
 189         '100': 'webm',
 190         '101': 'webm',
 191         '102': 'webm',
 192
 193         # videos that use m3u8
 194         '92': 'mp4',
 195         '93': 'mp4',
 196         '94': 'mp4',
 197         '95': 'mp4',
 198         '96': 'mp4',
 199         '132': 'mp4',
 200         '151': 'mp4',
 201
 202         # Dash mp4
 203         '133': 'mp4',
 204         '134': 'mp4',
 205         '135': 'mp4',
 206         '136': 'mp4',
 207         '137': 'mp4',
 208         '138': 'mp4',
 209         '139': 'mp4',
 210         '140': 'mp4',
 211         '141': 'mp4',
 212         '160': 'mp4',
 213
 214         # Dash webm
 215         '171': 'webm',
 216         '172': 'webm',
 217         '242': 'webm',
 218         '243': 'webm',
 219         '244': 'webm',
 220         '245': 'webm',
 221         '246': 'webm',
 222         '247': 'webm',
 223         '248': 'webm',
 224     }
 225     _video_dimensions = {
 226         '5': '240x400',
 227         '6': '???',
 228         '13': '???',
 229         '17': '144x176',
 230         '18': '360x640',
 231         '22': '720x1280',
 232         '34': '360x640',
 233         '35': '480x854',
 234         '37': '1080x1920',
 235         '38': '3072x4096',
 236         '43': '360x640',
 237         '44': '480x854',
 238         '45': '720x1280',
 239         '46': '1080x1920',
 240         '82': '360p',
 241         '83': '480p',
 242         '84': '720p',
 243         '85': '1080p',
 244         '92': '240p',
 245         '93': '360p',
 246         '94': '480p',
 247         '95': '720p',
 248         '96': '1080p',
 249         '100': '360p',
 250         '101': '480p',
 251         '102': '720p',
 252         '132': '240p',
 253         '151': '72p',
 254         '133': '240p',
 255         '134': '360p',
 256         '135': '480p',
 257         '136': '720p',
 258         '137': '1080p',
 259         '138': '>1080p',
 260         '139': '48k',
 261         '140': '128k',
 262         '141': '256k',
 263         '160': '192p',
 264         '171': '128k',
 265         '172': '256k',
 266         '242': '240p',
 267         '243': '360p',
 268         '244': '480p',
 269         '245': '480p',
 270         '246': '480p',
 271         '247': '720p',
 272         '248': '1080p',
 273     }
 274     _special_itags = {
 275         '82': '3D',
 276         '83': '3D',
 277         '84': '3D',
 278         '85': '3D',
 279         '100': '3D',
 280         '101': '3D',
 281         '102': '3D',
 282         '133': 'DASH Video',
 283         '134': 'DASH Video',
 284         '135': 'DASH Video',
 285         '136': 'DASH Video',
 286         '137': 'DASH Video',
 287         '138': 'DASH Video',
 288         '139': 'DASH Audio',
 289         '140': 'DASH Audio',
 290         '141': 'DASH Audio',
 291         '160': 'DASH Video',
 292         '171': 'DASH Audio',
 293         '172': 'DASH Audio',
 294         '242': 'DASH Video',
 295         '243': 'DASH Video',
 296         '244': 'DASH Video',
 297         '245': 'DASH Video',
 298         '246': 'DASH Video',
 299         '247': 'DASH Video',
 300         '248': 'DASH Video',
 301     }
 302
 303     IE_NAME = u'youtube'
 304     _TESTS = [
 305         {
 306             u"url":  u"http://www.youtube.com/watch?v=BaW_jenozKc",
 307             u"file":  u"BaW_jenozKc.mp4",
 308             u"info_dict": {
 309                 u"title": u"youtube-dl test video \"'/\\ä↭𝕐",
 310                 u"uploader": u"Philipp Hagemeister",
 311                 u"uploader_id": u"phihag",
 312                 u"upload_date": u"20121002",
 313                 u"description": u"test chars:  \"'/\\ä↭𝕐\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ."
 314             }
 315         },
 316         {
 317             u"url":  u"http://www.youtube.com/watch?v=1ltcDfZMA3U",
 318             u"file":  u"1ltcDfZMA3U.flv",
 319             u"note": u"Test VEVO video (#897)",
 320             u"info_dict": {
 321                 u"upload_date": u"20070518",
 322                 u"title": u"Maps - It Will Find You",
 323                 u"description": u"Music video by Maps performing It Will Find You.",
 324                 u"uploader": u"MuteUSA",
 325                 u"uploader_id": u"MuteUSA"
 326             }
 327         },
 328         {
 329             u"url":  u"http://www.youtube.com/watch?v=UxxajLWwzqY",
 330             u"file":  u"UxxajLWwzqY.mp4",
 331             u"note": u"Test generic use_cipher_signature video (#897)",
 332             u"info_dict": {
 333                 u"upload_date": u"20120506",
 334                 u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]",
 335                 u"description": u"md5:b085c9804f5ab69f4adea963a2dceb3c",
 336                 u"uploader": u"IconaPop",
 337                 u"uploader_id": u"IconaPop"
 338             }
 339         },
 340         {
 341             u"url":  u"https://www.youtube.com/watch?v=07FYdnEawAQ",
 342             u"file":  u"07FYdnEawAQ.mp4",
 343             u"note": u"Test VEVO video with age protection (#956)",
 344             u"info_dict": {
 345                 u"upload_date": u"20130703",
 346                 u"title": u"Justin Timberlake - Tunnel Vision (Explicit)",
 347                 u"description": u"md5:64249768eec3bc4276236606ea996373",
 348                 u"uploader": u"justintimberlakeVEVO",
 349                 u"uploader_id": u"justintimberlakeVEVO"
 350             }
 351         },
 352         {
 353             u'url': u'https://www.youtube.com/watch?v=TGi3HqYrWHE',
 354             u'file': u'TGi3HqYrWHE.mp4',
 355             u'note': u'm3u8 video',
 356             u'info_dict': {
 357                 u'title': u'Triathlon - Men - London 2012 Olympic Games',
 358                 u'description': u'- Men -  TR02 - Triathlon - 07 August 2012 - London 2012 Olympic Games',
 359                 u'uploader': u'olympic',
 360                 u'upload_date': u'20120807',
 361                 u'uploader_id': u'olympic',
 362             },
 363             u'params': {
 364                 u'skip_download': True,
 365             },
 366         },
 367     ]
 368
 369
 370     @classmethod
 371     def suitable(cls, url):
 372         """Receives a URL and returns True if suitable for this IE."""
 373         if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False
 374         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 375
 376     def report_video_webpage_download(self, video_id):
 377         """Report attempt to download video webpage."""
 378         self.to_screen(u'%s: Downloading video webpage' % video_id)
 379
 380     def report_video_info_webpage_download(self, video_id):
 381         """Report attempt to download video info webpage."""
 382         self.to_screen(u'%s: Downloading video info webpage' % video_id)
 383
 384     def report_video_subtitles_download(self, video_id):
 385         """Report attempt to download video info webpage."""
 386         self.to_screen(u'%s: Checking available subtitles' % video_id)
 387
 388     def report_video_subtitles_request(self, video_id, sub_lang, format):
 389         """Report attempt to download video info webpage."""
 390         self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format))
 391
 392     def report_video_subtitles_available(self, video_id, sub_lang_list):
 393         """Report available subtitles."""
 394         sub_lang = ",".join(list(sub_lang_list.keys()))
 395         self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang))
 396
 397     def report_information_extraction(self, video_id):
 398         """Report attempt to extract video information."""
 399         self.to_screen(u'%s: Extracting video information' % video_id)
 400
 401     def report_unavailable_format(self, video_id, format):
 402         """Report extracted video URL."""
 403         self.to_screen(u'%s: Format %s not available' % (video_id, format))
 404
 405     def report_rtmp_download(self):
 406         """Indicate the download will use the RTMP protocol."""
 407         self.to_screen(u'RTMP download detected')
 408
 409     def _decrypt_signature(self, s):
 410         """Turn the encrypted s field into a working signature"""
 411
 412         if len(s) == 92:
 413             return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
 414         elif len(s) == 90:
 415             return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81]
 416         elif len(s) == 89:
 417             return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1]
 418         elif len(s) == 88:
 419             return s[48] + s[81:67:-1] + s[82] + s[66:62:-1] + s[85] + s[61:48:-1] + s[67] + s[47:12:-1] + s[3] + s[11:3:-1] + s[2] + s[12]
 420         elif len(s) == 87:
 421             return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:]
 422         elif len(s) == 86:
 423             return s[5:20] + s[2] + s[21:]
 424         elif len(s) == 85:
 425             return s[83:34:-1] + s[0] + s[33:27:-1] + s[3] + s[26:19:-1] + s[34] + s[18:3:-1] + s[27]
 426         elif len(s) == 84:
 427             return s[83:27:-1] + s[0] + s[26:5:-1] + s[2:0:-1] + s[27]
 428         elif len(s) == 83:
 429             return s[81:64:-1] + s[82] + s[63:52:-1] + s[45] + s[51:45:-1] + s[1] + s[44:1:-1] + s[0]
 430         elif len(s) == 82:
 431             return s[36] + s[79:67:-1] + s[81] + s[66:40:-1] + s[33] + s[39:36:-1] + s[40] + s[35] + s[0] + s[67] + s[32:0:-1] + s[34]
 432         elif len(s) == 81:
 433             return s[56] + s[79:56:-1] + s[41] + s[55:41:-1] + s[80] + s[40:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
 434         elif len(s) == 79:
 435             return s[54] + s[77:54:-1] + s[39] + s[53:39:-1] + s[78] + s[38:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
 436
 437         else:
 438             raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s)))
 439
 440     def _decrypt_signature_age_gate(self, s):
 441         # The videos with age protection use another player, so the algorithms
 442         # can be different.
 443         if len(s) == 86:
 444             return s[2:63] + s[82] + s[64:82] + s[63]
 445         else:
 446             # Fallback to the other algortihms
 447             return self._decrypt_signature(s)
 448
 449
 450     def _get_available_subtitles(self, video_id):
 451         self.report_video_subtitles_download(video_id)
 452         request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
 453         try:
 454             sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
 455         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 456             return (u'unable to download video subtitles: %s' % compat_str(err), None)
 457         sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
 458         sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list)
 459         if not sub_lang_list:
 460             return (u'video doesn\'t have subtitles', None)
 461         return sub_lang_list
 462
 463     def _list_available_subtitles(self, video_id):
 464         sub_lang_list = self._get_available_subtitles(video_id)
 465         self.report_video_subtitles_available(video_id, sub_lang_list)
 466
 467     def _request_subtitle(self, sub_lang, sub_name, video_id, format):
 468         """
 469         Return tuple:
 470         (error_message, sub_lang, sub)
 471         """
 472         self.report_video_subtitles_request(video_id, sub_lang, format)
 473         params = compat_urllib_parse.urlencode({
 474             'lang': sub_lang,
 475             'name': sub_name,
 476             'v': video_id,
 477             'fmt': format,
 478         })
 479         url = 'http://www.youtube.com/api/timedtext?' + params
 480         try:
 481             sub = compat_urllib_request.urlopen(url).read().decode('utf-8')
 482         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 483             return (u'unable to download video subtitles: %s' % compat_str(err), None, None)
 484         if not sub:
 485             return (u'Did not fetch video subtitles', None, None)
 486         return (None, sub_lang, sub)
 487
 488     def _request_automatic_caption(self, video_id, webpage):
 489         """We need the webpage for getting the captions url, pass it as an
 490            argument to speed up the process."""
 491         sub_lang = self._downloader.params.get('subtitleslang') or 'en'
 492         sub_format = self._downloader.params.get('subtitlesformat')
 493         self.to_screen(u'%s: Looking for automatic captions' % video_id)
 494         mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
 495         err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
 496         if mobj is None:
 497             return [(err_msg, None, None)]
 498         player_config = json.loads(mobj.group(1))
 499         try:
 500             args = player_config[u'args']
 501             caption_url = args[u'ttsurl']
 502             timestamp = args[u'timestamp']
 503             params = compat_urllib_parse.urlencode({
 504                 'lang': 'en',
 505                 'tlang': sub_lang,
 506                 'fmt': sub_format,
 507                 'ts': timestamp,
 508                 'kind': 'asr',
 509             })
 510             subtitles_url = caption_url + '&' + params
 511             sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
 512             return [(None, sub_lang, sub)]
 513         except KeyError:
 514             return [(err_msg, None, None)]
 515
 516     def _extract_subtitle(self, video_id):
 517         """
 518         Return a list with a tuple:
 519         [(error_message, sub_lang, sub)]
 520         """
 521         sub_lang_list = self._get_available_subtitles(video_id)
 522         sub_format = self._downloader.params.get('subtitlesformat')
 523         if  isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
 524             return [(sub_lang_list[0], None, None)]
 525         if self._downloader.params.get('subtitleslang', False):
 526             sub_lang = self._downloader.params.get('subtitleslang')
 527         elif 'en' in sub_lang_list:
 528             sub_lang = 'en'
 529         else:
 530             sub_lang = list(sub_lang_list.keys())[0]
 531         if not sub_lang in sub_lang_list:
 532             return [(u'no closed captions found in the specified language "%s"' % sub_lang, None, None)]
 533
 534         subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 535         return [subtitle]
 536
 537     def _extract_all_subtitles(self, video_id):
 538         sub_lang_list = self._get_available_subtitles(video_id)
 539         sub_format = self._downloader.params.get('subtitlesformat')
 540         if  isinstance(sub_lang_list,tuple): #There was some error, it didn't get the available subtitles
 541             return [(sub_lang_list[0], None, None)]
 542         subtitles = []
 543         for sub_lang in sub_lang_list:
 544             subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format)
 545             subtitles.append(subtitle)
 546         return subtitles
 547
 548     def _print_formats(self, formats):
 549         print('Available formats:')
 550         for x in formats:
 551             print('%s\t:\t%s\t[%s]%s' %(x, self._video_extensions.get(x, 'flv'),
 552                                         self._video_dimensions.get(x, '???'),
 553                                         ' ('+self._special_itags[x]+')' if x in self._special_itags else ''))
 554
 555     def _extract_id(self, url):
 556         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 557         if mobj is None:
 558             raise ExtractorError(u'Invalid URL: %s' % url)
 559         video_id = mobj.group(2)
 560         return video_id
 561
 562     def _get_video_url_list(self, url_map):
 563         """
 564         Transform a dictionary in the format {itag:url} to a list of (itag, url)
 565         with the requested formats.
 566         """
 567         req_format = self._downloader.params.get('format', None)
 568         format_limit = self._downloader.params.get('format_limit', None)
 569         available_formats = self._available_formats_prefer_free if self._downloader.params.get('prefer_free_formats', False) else self._available_formats
 570         if format_limit is not None and format_limit in available_formats:
 571             format_list = available_formats[available_formats.index(format_limit):]
 572         else:
 573             format_list = available_formats
 574         existing_formats = [x for x in format_list if x in url_map]
 575         if len(existing_formats) == 0:
 576             raise ExtractorError(u'no known formats available for video')
 577         if self._downloader.params.get('listformats', None):
 578             self._print_formats(existing_formats)
 579             return
 580         if req_format is None or req_format == 'best':
 581             video_url_list = [(existing_formats[0], url_map[existing_formats[0]])] # Best quality
 582         elif req_format == 'worst':
 583             video_url_list = [(existing_formats[-1], url_map[existing_formats[-1]])] # worst quality
 584         elif req_format in ('-1', 'all'):
 585             video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
 586         else:
 587             # Specific formats. We pick the first in a slash-delimeted sequence.
 588             # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'.
 589             req_formats = req_format.split('/')
 590             video_url_list = None
 591             for rf in req_formats:
 592                 if rf in url_map:
 593                     video_url_list = [(rf, url_map[rf])]
 594                     break
 595             if video_url_list is None:
 596                 raise ExtractorError(u'requested format not available')
 597         return video_url_list
 598
 599     def _extract_from_m3u8(self, manifest_url, video_id):
 600         url_map = {}
 601         def _get_urls(_manifest):
 602             lines = _manifest.split('\n')
 603             urls = filter(lambda l: l and not l.startswith('#'),
 604                             lines)
 605             return urls
 606         manifest = self._download_webpage(manifest_url, video_id, u'Downloading formats manifest')
 607         formats_urls = _get_urls(manifest)
 608         for format_url in formats_urls:
 609             itag = self._search_regex(r'itag/(\d+?)/', format_url, 'itag')
 610             url_map[itag] = format_url
 611         return url_map
 612
 613     def _real_extract(self, url):
 614         if re.match(r'(?:https?://)?[^/]+/watch\?feature=[a-z_]+$', url):
 615             self._downloader.report_warning(u'Did you forget to quote the URL? Remember that & is a meta-character in most shells, so you want to put the URL in quotes, like  youtube-dl \'http://www.youtube.com/watch?feature=foo&v=BaW_jenozKc\' (or simply  youtube-dl BaW_jenozKc  ).')
 616
 617         # Extract original video URL from URL with redirection, like age verification, using next_url parameter
 618         mobj = re.search(self._NEXT_URL_RE, url)
 619         if mobj:
 620             url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/')
 621         video_id = self._extract_id(url)
 622
 623         # Get video webpage
 624         self.report_video_webpage_download(video_id)
 625         url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id
 626         request = compat_urllib_request.Request(url)
 627         try:
 628             video_webpage_bytes = compat_urllib_request.urlopen(request).read()
 629         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 630             raise ExtractorError(u'Unable to download video webpage: %s' % compat_str(err))
 631
 632         video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
 633
 634         # Attempt to extract SWF player URL
 635         mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
 636         if mobj is not None:
 637             player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
 638         else:
 639             player_url = None
 640
 641         # Get video info
 642         self.report_video_info_webpage_download(video_id)
 643         if re.search(r'player-age-gate-content">', video_webpage) is not None:
 644             self.report_age_confirmation()
 645             age_gate = True
 646             # We simulate the access to the video from www.youtube.com/v/{video_id}
 647             # this can be viewed without login into Youtube
 648             data = compat_urllib_parse.urlencode({'video_id': video_id,
 649                                                   'el': 'embedded',
 650                                                   'gl': 'US',
 651                                                   'hl': 'en',
 652                                                   'eurl': 'https://youtube.googleapis.com/v/' + video_id,
 653                                                   'asv': 3,
 654                                                   'sts':'1588',
 655                                                   })
 656             video_info_url = 'https://www.youtube.com/get_video_info?' + data
 657             video_info_webpage = self._download_webpage(video_info_url, video_id,
 658                                     note=False,
 659                                     errnote='unable to download video info webpage')
 660             video_info = compat_parse_qs(video_info_webpage)
 661         else:
 662             age_gate = False
 663             for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
 664                 video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
 665                         % (video_id, el_type))
 666                 video_info_webpage = self._download_webpage(video_info_url, video_id,
 667                                         note=False,
 668                                         errnote='unable to download video info webpage')
 669                 video_info = compat_parse_qs(video_info_webpage)
 670                 if 'token' in video_info:
 671                     break
 672         if 'token' not in video_info:
 673             if 'reason' in video_info:
 674                 raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True)
 675             else:
 676                 raise ExtractorError(u'"token" parameter not in video info for unknown reason')
 677
 678         # Check for "rental" videos
 679         if 'ypc_video_rental_bar_text' in video_info and 'author' not in video_info:
 680             raise ExtractorError(u'"rental" videos not supported')
 681
 682         # Start extracting information
 683         self.report_information_extraction(video_id)
 684
 685         # uploader
 686         if 'author' not in video_info:
 687             raise ExtractorError(u'Unable to extract uploader name')
 688         video_uploader = compat_urllib_parse.unquote_plus(video_info['author'][0])
 689
 690         # uploader_id
 691         video_uploader_id = None
 692         mobj = re.search(r'<link itemprop="url" href="http://www.youtube.com/(?:user|channel)/([^"]+)">', video_webpage)
 693         if mobj is not None:
 694             video_uploader_id = mobj.group(1)
 695         else:
 696             self._downloader.report_warning(u'unable to extract uploader nickname')
 697
 698         # title
 699         if 'title' not in video_info:
 700             raise ExtractorError(u'Unable to extract video title')
 701         video_title = compat_urllib_parse.unquote_plus(video_info['title'][0])
 702
 703         # thumbnail image
 704         # We try first to get a high quality image:
 705         m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">',
 706                             video_webpage, re.DOTALL)
 707         if m_thumb is not None:
 708             video_thumbnail = m_thumb.group(1)
 709         elif 'thumbnail_url' not in video_info:
 710             self._downloader.report_warning(u'unable to extract video thumbnail')
 711             video_thumbnail = ''
 712         else:   # don't panic if we can't find it
 713             video_thumbnail = compat_urllib_parse.unquote_plus(video_info['thumbnail_url'][0])
 714
 715         # upload date
 716         upload_date = None
 717         mobj = re.search(r'id="eow-date.*?>(.*?)</span>', video_webpage, re.DOTALL)
 718         if mobj is not None:
 719             upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split())
 720             upload_date = unified_strdate(upload_date)
 721
 722         # description
 723         video_description = get_element_by_id("eow-description", video_webpage)
 724         if video_description:
 725             video_description = clean_html(video_description)
 726         else:
 727             fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
 728             if fd_mobj:
 729                 video_description = unescapeHTML(fd_mobj.group(1))
 730             else:
 731                 video_description = u''
 732
 733         # subtitles
 734         video_subtitles = None
 735
 736         if self._downloader.params.get('writesubtitles', False):
 737             video_subtitles = self._extract_subtitle(video_id)
 738             if video_subtitles:
 739                 (sub_error, sub_lang, sub) = video_subtitles[0]
 740                 if sub_error:
 741                     self._downloader.report_warning(sub_error)
 742
 743         if self._downloader.params.get('writeautomaticsub', False):
 744             video_subtitles = self._request_automatic_caption(video_id, video_webpage)
 745             (sub_error, sub_lang, sub) = video_subtitles[0]
 746             if sub_error:
 747                 self._downloader.report_warning(sub_error)
 748
 749         if self._downloader.params.get('allsubtitles', False):
 750             video_subtitles = self._extract_all_subtitles(video_id)
 751             for video_subtitle in video_subtitles:
 752                 (sub_error, sub_lang, sub) = video_subtitle
 753                 if sub_error:
 754                     self._downloader.report_warning(sub_error)
 755
 756         if self._downloader.params.get('listsubtitles', False):
 757             self._list_available_subtitles(video_id)
 758             return
 759
 760         if 'length_seconds' not in video_info:
 761             self._downloader.report_warning(u'unable to extract video duration')
 762             video_duration = ''
 763         else:
 764             video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0])
 765
 766         # Decide which formats to download
 767
 768         try:
 769             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
 770             if not mobj:
 771                 raise ValueError('Could not find vevo ID')
 772             info = json.loads(mobj.group(1))
 773             args = info['args']
 774             # Easy way to know if the 's' value is in url_encoded_fmt_stream_map
 775             # this signatures are encrypted
 776             m_s = re.search(r'[&,]s=', args['url_encoded_fmt_stream_map'])
 777             if m_s is not None:
 778                 self.to_screen(u'%s: Encrypted signatures detected.' % video_id)
 779                 video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
 780             m_s = re.search(r'[&,]s=', args.get('adaptive_fmts', u''))
 781             if m_s is not None:
 782                 if 'url_encoded_fmt_stream_map' in video_info:
 783                     video_info['url_encoded_fmt_stream_map'][0] += ',' + args['adaptive_fmts']
 784                 else:
 785                     video_info['url_encoded_fmt_stream_map'] = [args['adaptive_fmts']]
 786             elif 'adaptive_fmts' in video_info:
 787                 if 'url_encoded_fmt_stream_map' in video_info:
 788                     video_info['url_encoded_fmt_stream_map'][0] += ',' + video_info['adaptive_fmts'][0]
 789                 else:
 790                     video_info['url_encoded_fmt_stream_map'] = video_info['adaptive_fmts']
 791         except ValueError:
 792             pass
 793
 794         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
 795             self.report_rtmp_download()
 796             video_url_list = [(None, video_info['conn'][0])]
 797         elif 'url_encoded_fmt_stream_map' in video_info and len(video_info['url_encoded_fmt_stream_map']) >= 1:
 798             if 'rtmpe%3Dyes' in video_info['url_encoded_fmt_stream_map'][0]:
 799                 raise ExtractorError('rtmpe downloads are not supported, see https://github.com/rg3/youtube-dl/issues/343 for more information.', expected=True)
 800             url_map = {}
 801             for url_data_str in video_info['url_encoded_fmt_stream_map'][0].split(','):
 802                 url_data = compat_parse_qs(url_data_str)
 803                 if 'itag' in url_data and 'url' in url_data:
 804                     url = url_data['url'][0]
 805                     if 'sig' in url_data:
 806                         url += '&signature=' + url_data['sig'][0]
 807                     elif 's' in url_data:
 808                         if self._downloader.params.get('verbose'):
 809                             s = url_data['s'][0]
 810                             if age_gate:
 811                                 player_version = self._search_regex(r'ad3-(.+?)\.swf',
 812                                     video_info['ad3_module'][0] if 'ad3_module' in video_info else 'NOT FOUND',
 813                                     'flash player', fatal=False)
 814                                 player = 'flash player %s' % player_version
 815                             else:
 816                                 player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
 817                                     'html5 player', fatal=False)
 818                             parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
 819                             self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
 820                                 (len(s), parts_sizes, url_data['itag'][0], player))
 821                         encrypted_sig = url_data['s'][0]
 822                         if age_gate:
 823                             signature = self._decrypt_signature_age_gate(encrypted_sig)
 824                         else:
 825                             signature = self._decrypt_signature(encrypted_sig)
 826                         url += '&signature=' + signature
 827                     if 'ratebypass' not in url:
 828                         url += '&ratebypass=yes'
 829                     url_map[url_data['itag'][0]] = url
 830             video_url_list = self._get_video_url_list(url_map)
 831             if not video_url_list:
 832                 return
 833         elif video_info.get('hlsvp'):
 834             manifest_url = video_info['hlsvp'][0]
 835             url_map = self._extract_from_m3u8(manifest_url, video_id)
 836             video_url_list = self._get_video_url_list(url_map)
 837             if not video_url_list:
 838                 return
 839
 840         else:
 841             raise ExtractorError(u'no conn or url_encoded_fmt_stream_map information found in video info')
 842
 843         results = []
 844         for format_param, video_real_url in video_url_list:
 845             # Extension
 846             video_extension = self._video_extensions.get(format_param, 'flv')
 847
 848             video_format = '{0} - {1}{2}'.format(format_param if format_param else video_extension,
 849                                               self._video_dimensions.get(format_param, '???'),
 850                                               ' ('+self._special_itags[format_param]+')' if format_param in self._special_itags else '')
 851
 852             results.append({
 853                 'id':       video_id,
 854                 'url':      video_real_url,
 855                 'uploader': video_uploader,
 856                 'uploader_id': video_uploader_id,
 857                 'upload_date':  upload_date,
 858                 'title':    video_title,
 859                 'ext':      video_extension,
 860                 'format':   video_format,
 861                 'thumbnail':    video_thumbnail,
 862                 'description':  video_description,
 863                 'player_url':   player_url,
 864                 'subtitles':    video_subtitles,
 865                 'duration':     video_duration
 866             })
 867         return results
 868
 869 class YoutubePlaylistIE(InfoExtractor):
 870     IE_DESC = u'YouTube.com playlists'
 871     _VALID_URL = r"""(?:
 872                         (?:https?://)?
 873                         (?:\w+\.)?
 874                         youtube\.com/
 875                         (?:
 876                            (?:course|view_play_list|my_playlists|artist|playlist|watch)
 877                            \? (?:.*?&)*? (?:p|a|list)=
 878                         |  p/
 879                         )
 880                         ((?:PL|EC|UU|FL)?[0-9A-Za-z-_]{10,})
 881                         .*
 882                      |
 883                         ((?:PL|EC|UU|FL)[0-9A-Za-z-_]{10,})
 884                      )"""
 885     _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/playlists/%s?max-results=%i&start-index=%i&v=2&alt=json&safeSearch=none'
 886     _MAX_RESULTS = 50
 887     IE_NAME = u'youtube:playlist'
 888
 889     @classmethod
 890     def suitable(cls, url):
 891         """Receives a URL and returns True if suitable for this IE."""
 892         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
 893
 894     def _real_extract(self, url):
 895         # Extract playlist id
 896         mobj = re.match(self._VALID_URL, url, re.VERBOSE)
 897         if mobj is None:
 898             raise ExtractorError(u'Invalid URL: %s' % url)
 899
 900         # Download playlist videos from API
 901         playlist_id = mobj.group(1) or mobj.group(2)
 902         videos = []
 903
 904         for page_num in itertools.count(1):
 905             start_index = self._MAX_RESULTS * (page_num - 1) + 1
 906             if start_index >= 1000:
 907                 self._downloader.report_warning(u'Max number of results reached')
 908                 break
 909             url = self._TEMPLATE_URL % (playlist_id, self._MAX_RESULTS, start_index)
 910             page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num)
 911
 912             try:
 913                 response = json.loads(page)
 914             except ValueError as err:
 915                 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
 916
 917             if 'feed' not in response:
 918                 raise ExtractorError(u'Got a malformed response from YouTube API')
 919             playlist_title = response['feed']['title']['$t']
 920             if 'entry' not in response['feed']:
 921                 # Number of videos is a multiple of self._MAX_RESULTS
 922                 break
 923
 924             for entry in response['feed']['entry']:
 925                 index = entry['yt$position']['$t']
 926                 if 'media$group' in entry and 'media$player' in entry['media$group']:
 927                     videos.append((index, entry['media$group']['media$player']['url']))
 928
 929         videos = [v[1] for v in sorted(videos)]
 930
 931         url_results = [self.url_result(vurl, 'Youtube') for vurl in videos]
 932         return [self.playlist_result(url_results, playlist_id, playlist_title)]
 933
 934
 935 class YoutubeChannelIE(InfoExtractor):
 936     IE_DESC = u'YouTube.com channels'
 937     _VALID_URL = r"^(?:https?://)?(?:youtu\.be|(?:\w+\.)?youtube(?:-nocookie)?\.com)/channel/([0-9A-Za-z_-]+)"
 938     _TEMPLATE_URL = 'http://www.youtube.com/channel/%s/videos?sort=da&flow=list&view=0&page=%s&gl=US&hl=en'
 939     _MORE_PAGES_INDICATOR = 'yt-uix-load-more'
 940     _MORE_PAGES_URL = 'http://www.youtube.com/c4_browse_ajax?action_load_more_videos=1&flow=list&paging=%s&view=0&sort=da&channel_id=%s'
 941     IE_NAME = u'youtube:channel'
 942
 943     def extract_videos_from_page(self, page):
 944         ids_in_page = []
 945         for mobj in re.finditer(r'href="/watch\?v=([0-9A-Za-z_-]+)&?', page):
 946             if mobj.group(1) not in ids_in_page:
 947                 ids_in_page.append(mobj.group(1))
 948         return ids_in_page
 949
 950     def _real_extract(self, url):
 951         # Extract channel id
 952         mobj = re.match(self._VALID_URL, url)
 953         if mobj is None:
 954             raise ExtractorError(u'Invalid URL: %s' % url)
 955
 956         # Download channel page
 957         channel_id = mobj.group(1)
 958         video_ids = []
 959         pagenum = 1
 960
 961         url = self._TEMPLATE_URL % (channel_id, pagenum)
 962         page = self._download_webpage(url, channel_id,
 963                                       u'Downloading page #%s' % pagenum)
 964
 965         # Extract video identifiers
 966         ids_in_page = self.extract_videos_from_page(page)
 967         video_ids.extend(ids_in_page)
 968
 969         # Download any subsequent channel pages using the json-based channel_ajax query
 970         if self._MORE_PAGES_INDICATOR in page:
 971             for pagenum in itertools.count(1):
 972                 url = self._MORE_PAGES_URL % (pagenum, channel_id)
 973                 page = self._download_webpage(url, channel_id,
 974                                               u'Downloading page #%s' % pagenum)
 975
 976                 page = json.loads(page)
 977
 978                 ids_in_page = self.extract_videos_from_page(page['content_html'])
 979                 video_ids.extend(ids_in_page)
 980
 981                 if self._MORE_PAGES_INDICATOR  not in page['load_more_widget_html']:
 982                     break
 983
 984         self._downloader.to_screen(u'[youtube] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
 985
 986         urls = ['http://www.youtube.com/watch?v=%s' % id for id in video_ids]
 987         url_entries = [self.url_result(eurl, 'Youtube') for eurl in urls]
 988         return [self.playlist_result(url_entries, channel_id)]
 989
 990
 991 class YoutubeUserIE(InfoExtractor):
 992     IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)'
 993     _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/user/)|ytuser:)([A-Za-z0-9_-]+)'
 994     _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s'
 995     _GDATA_PAGE_SIZE = 50
 996     _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d'
 997     _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]'
 998     IE_NAME = u'youtube:user'
 999
1000     def _real_extract(self, url):
1001         # Extract username
1002         mobj = re.match(self._VALID_URL, url)
1003         if mobj is None:
1004             raise ExtractorError(u'Invalid URL: %s' % url)
1005
1006         username = mobj.group(1)
1007
1008         # Download video ids using YouTube Data API. Result size per
1009         # query is limited (currently to 50 videos) so we need to query
1010         # page by page until there are no video ids - it means we got
1011         # all of them.
1012
1013         video_ids = []
1014
1015         for pagenum in itertools.count(0):
1016             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
1017
1018             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
1019             page = self._download_webpage(gdata_url, username,
1020                                           u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
1021
1022             # Extract video identifiers
1023             ids_in_page = []
1024
1025             for mobj in re.finditer(self._VIDEO_INDICATOR, page):
1026                 if mobj.group(1) not in ids_in_page:
1027                     ids_in_page.append(mobj.group(1))
1028
1029             video_ids.extend(ids_in_page)
1030
1031             # A little optimization - if current page is not
1032             # "full", ie. does not contain PAGE_SIZE video ids then
1033             # we can assume that this page is the last one - there
1034             # are no more ids on further pages - no need to query
1035             # again.
1036
1037             if len(ids_in_page) < self._GDATA_PAGE_SIZE:
1038                 break
1039
1040         urls = ['http://www.youtube.com/watch?v=%s' % video_id for video_id in video_ids]
1041         url_results = [self.url_result(rurl, 'Youtube') for rurl in urls]
1042         return [self.playlist_result(url_results, playlist_title = username)]
1043
1044 class YoutubeSearchIE(SearchInfoExtractor):
1045     IE_DESC = u'YouTube.com searches'
1046     _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc'
1047     _MAX_RESULTS = 1000
1048     IE_NAME = u'youtube:search'
1049     _SEARCH_KEY = 'ytsearch'
1050
1051     def report_download_page(self, query, pagenum):
1052         """Report attempt to download search page with given number."""
1053         self._downloader.to_screen(u'[youtube] query "%s": Downloading page %s' % (query, pagenum))
1054
1055     def _get_n_results(self, query, n):
1056         """Get a specified number of results for a query"""
1057
1058         video_ids = []
1059         pagenum = 0
1060         limit = n
1061
1062         while (50 * pagenum) < limit:
1063             self.report_download_page(query, pagenum+1)
1064             result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1)
1065             request = compat_urllib_request.Request(result_url)
1066             try:
1067                 data = compat_urllib_request.urlopen(request).read().decode('utf-8')
1068             except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1069                 raise ExtractorError(u'Unable to download API page: %s' % compat_str(err))
1070             api_response = json.loads(data)['data']
1071
1072             if not 'items' in api_response:
1073                 raise ExtractorError(u'[youtube] No video results')
1074
1075             new_ids = list(video['id'] for video in api_response['items'])
1076             video_ids += new_ids
1077
1078             limit = min(n, api_response['totalItems'])
1079             pagenum += 1
1080
1081         if len(video_ids) > n:
1082             video_ids = video_ids[:n]
1083         videos = [self.url_result('http://www.youtube.com/watch?v=%s' % id, 'Youtube') for id in video_ids]
1084         return self.playlist_result(videos, query)
1085
1086
1087 class YoutubeShowIE(InfoExtractor):
1088     IE_DESC = u'YouTube.com (multi-season) shows'
1089     _VALID_URL = r'https?://www\.youtube\.com/show/(.*)'
1090     IE_NAME = u'youtube:show'
1091
1092     def _real_extract(self, url):
1093         mobj = re.match(self._VALID_URL, url)
1094         show_name = mobj.group(1)
1095         webpage = self._download_webpage(url, show_name, u'Downloading show webpage')
1096         # There's one playlist for each season of the show
1097         m_seasons = list(re.finditer(r'href="(/playlist\?list=.*?)"', webpage))
1098         self.to_screen(u'%s: Found %s seasons' % (show_name, len(m_seasons)))
1099         return [self.url_result('https://www.youtube.com' + season.group(1), 'YoutubePlaylist') for season in m_seasons]
1100
1101
1102 class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor):
1103     """
1104     Base class for extractors that fetch info from
1105     http://www.youtube.com/feed_ajax
1106     Subclasses must define the _FEED_NAME and _PLAYLIST_TITLE properties.
1107     """
1108     _LOGIN_REQUIRED = True
1109     _PAGING_STEP = 30
1110     # use action_load_personal_feed instead of action_load_system_feed
1111     _PERSONAL_FEED = False
1112
1113     @property
1114     def _FEED_TEMPLATE(self):
1115         action = 'action_load_system_feed'
1116         if self._PERSONAL_FEED:
1117             action = 'action_load_personal_feed'
1118         return 'http://www.youtube.com/feed_ajax?%s=1&feed_name=%s&paging=%%s' % (action, self._FEED_NAME)
1119
1120     @property
1121     def IE_NAME(self):
1122         return u'youtube:%s' % self._FEED_NAME
1123
1124     def _real_initialize(self):
1125         self._login()
1126
1127     def _real_extract(self, url):
1128         feed_entries = []
1129         # The step argument is available only in 2.7 or higher
1130         for i in itertools.count(0):
1131             paging = i*self._PAGING_STEP
1132             info = self._download_webpage(self._FEED_TEMPLATE % paging,
1133                                           u'%s feed' % self._FEED_NAME,
1134                                           u'Downloading page %s' % i)
1135             info = json.loads(info)
1136             feed_html = info['feed_html']
1137             m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html)
1138             ids = orderedSet(m.group(1) for m in m_ids)
1139             feed_entries.extend(self.url_result(id, 'Youtube') for id in ids)
1140             if info['paging'] is None:
1141                 break
1142         return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE)
1143
1144 class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor):
1145     IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)'
1146     _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?'
1147     _FEED_NAME = 'subscriptions'
1148     _PLAYLIST_TITLE = u'Youtube Subscriptions'
1149
1150 class YoutubeRecommendedIE(YoutubeFeedsInfoExtractor):
1151     IE_DESC = u'YouTube.com recommended videos, "ytrec" keyword (requires authentication)'
1152     _VALID_URL = r'https?://www\.youtube\.com/feed/recommended|:ytrec(?:ommended)?'
1153     _FEED_NAME = 'recommended'
1154     _PLAYLIST_TITLE = u'Youtube Recommended videos'
1155
1156 class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor):
1157     IE_DESC = u'Youtube watch later list, "ytwatchlater" keyword (requires authentication)'
1158     _VALID_URL = r'https?://www\.youtube\.com/feed/watch_later|:ytwatchlater'
1159     _FEED_NAME = 'watch_later'
1160     _PLAYLIST_TITLE = u'Youtube Watch Later'
1161     _PAGING_STEP = 100
1162     _PERSONAL_FEED = True
1163
1164 class YoutubeFavouritesIE(YoutubeBaseInfoExtractor):
1165     IE_NAME = u'youtube:favorites'
1166     IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)'
1167     _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:o?rites)?'
1168     _LOGIN_REQUIRED = True
1169
1170     def _real_extract(self, url):
1171         webpage = self._download_webpage('https://www.youtube.com/my_favorites', 'Youtube Favourites videos')
1172         playlist_id = self._search_regex(r'list=(.+?)["&]', webpage, u'favourites playlist id')
1173         return self.url_result(playlist_id, 'YoutubePlaylist')